Evaluating the Correlation and Responsiveness of Patient-Reported Pain With Function and Quality-of-Life Outcomes After Spine Surgery

Study Design. Systematic review. Objective. To determine the correlation of patient-reported pain with physical function and health-related quality of life (HRQoL) after spine surgery and to determine the responsiveness of pain, physical function, and HRQoL after spine surgery. Summary of Background Data. Several validated outcome instruments are available to assess the success of treatment for chronic low back pain. These patient-centered tools include measurements for pain based on numeric scales, validated condition-specific functional outcomes measures, and HRQoL outcomes measures. It is unclear whether these three types of patient-reported outcomes are measuring different constructs and whether all three should be measured after spine surgery. In addition, it is unclear which of these outcomes measures is most sensitive to change after spine surgery for low back pain. Methods. A systematic search was conducted in MEDLINE, EMBASE, and the Cochrane Collaboration Library for literature published through December 2010. The correlation between pain (visual analog scale, VAS), physical function (Oswestry Disability Index, ODI), and HRQoL (36-Item Short Form Health Survey [SF-36] and European Quality of Life [EQ-5D]) change scores was performed using the Spearman rank correlation coefficients. To compare the responsiveness of pain, function, and HRQoL scores after spine surgery, we calculated effect sizes by dividing change scores by the SD of the baseline scores. This standardized method allowed us to compare the responsiveness of each outcome measure directly and reported an effect size of 0.2 to 0.3 as a "small" effect, around 0.5 a "medium" effect and 0.8 to infinity, a "large" effect. To determine whether the differences in effect sizes measuring responsiveness were significantly different, we conducted a Wilcoxon signed-rank test between each of the three measurements of pain, function, and HRQoL scores when there was enough data to perform the test. Results. None of the correlations exceeded 0.70 using the Spearman rank correlation coefcients, suggesting that these outcomes are measuring different constructs. The strongest correlations were between the VAS back pain change scores and the SF-36 physical composite score change scores (rho = 0.67) and VAS back pain change scores and ODI change scores (rho = 0.69). The pooled mean effect sizes for the five studies that reported a pain measure and the ODI were 1.4 +/- 0.57 and 1.1 +/- 0.39, respectively. Both are considered "large" effect sizes. The pooled mean effect sizes for the three studies reporting the SF-36 physical and mental composite scores were 0.66 +/- 0.39 and 0.54 +/- 0.36, respectively. Both are considered "medium" effect sizes. The pooled mean effect sizes for the single studies reporting the EQ-5D and SF-36 total score were 0.78 +/- 0.12 and 0.34 +/- 0.21. These were "medium" and "small," respectively. Conclusion. We observed little correlation between the change in pain and the change in HRQoL outcomes measures. The strongest correlation was between VAS pain and ODI but was still not considered strong (0.69). These findings suggest that these three outcomes (pain, function, and HRQoL) are measuring different constructs. With respect to responsiveness, VAS pain and ODI were the only outcomes measures that demonstrated a large effect after lumbar spine surgery. None of the HRQoL tools were as sensitive to the treatment. The EQ-5D, SF physical composite, and SF mental composite outcomes demonstrated a medium effect, while the SF-36 total score demonstrated a small effect. The responsive measure shows that the more specific the outcomes tool, the more sensitive the response. Clinical Recommendations. Recommendation 1: When surgically treating CLBP, we recommend administering both a VAS for pain and a condition-specific physical measure such as the ODI before and after surgical intervention as these outcomes are the most treatment specific and responsive to change. Strength of recommendation: Strong. Recommendation 2: When evaluating the surgical outcomes for CLBP in the clinical-research setting, we recommend selecting a shorter version for measuring general HRQoL (e.g., SF-12, EQ-5D) to minimize clinician and patient burden. Strength of recommendation: Strong.